Sunday, January 2, 2011



By Bernie Bierman
One positive thing that can be said of the American Translators Association (ATA) is that it has a propaganda machine that works 24/7.  In fact, I would say that the ATA has a propaganda machine that is by far the best of all of the propaganda machines of all of the translator-interpreter organizations in the world.  Of course, one must always remember that propaganda and the dissemination thereof is what trade associations do best. In fact, one can say with a good deal of cogency that propaganda is the very raison d’être of a trade association.

But there are a couple of problems with propaganda, especially pristine pure propaganda.  One is that propaganda (and propagandized information) has (have) little connection to reality, and the other is that propaganda (and propagandized information) can place the members of a trade or profession in a light that isn’t too complimentary.  Dangerous stuff this propaganda.  Mishandling may cause injury.

“U.S. News & World Report” is a weekly news magazine published in the United States (where else?).  In its hey-day, it held third-place behind “Time” and “Newsweek” among the weekly news magazines.  With 24-hour cable TV news , which is also disseminated over the internet and a more than a discernible decline in reading (especially reading more than a few hundred characters a la Twitter), the three above-mentioned magazines along with their daily newspaper brethren have become almost creaking relics of a journalism past. 

On the eve before the eve of the new year, the propaganda machine of the American Translators Association loudly proclaimed to the ca. 10,000 faithful that translating and interpreting “were among the 50 best occupations projected for 2011”.  Do you think I’m joking?  Well, the proof is not only in the pudding, it is equally in the eating.  So here, have some ATA pudding as served up as an end-of-the-year dessert:

“U.S. News & World Report has rated translation and interpreting as one of the 50 best careers to pursue in 2011. The magazine’s annual review of high-opportunity professions has become a go-to resource for school guidance counselors, college students, job-seekers, and anyone who is considering a career change.

To come up with this year's top jobs list,
U.S. News considered job-growth projections, salary data, educational requirements, and market demand. Labor and industry experts also weighed in with stats on employment prospects and job satisfaction.”

Now then, folks, I would like you to focus in on that sentence “Labor and industry experts also weighed in with stats* on employment prospects and job satisfaction”.

* For those of you for whom English is a second language, please do not take it as an act of condescension or patronization if I explain to you that millions upon millions of native speakers of the language are afflicted with what may be called aggressively-malignant linguistic laziness, to wit:  “Congratulations” is a word said only by those born before 1955.  The word in 21st century English is “congrats”.  The only time that the word “application” is used is when one speaks of a job application.  What is used in your computer or iPod or iPad or mobile phone are “apps”.  And those “abdominal muscles” that so many of us are trying to keep firm are called “abs”.  And all those figures that we see in connection with all sorts of endeavors are called “stats”, since no native speaker of English would ever say “statistics”, except those doddering, incontinent soon-to-be-dead citizens.

And in the case of ATA, this is how these industry experts weighed in:

“Salary varies greatly depending on language and subject matter. Interpreters and translators who speak languages that are in high demand or underrepresented in the field often have higher earnings, as do those who communicate about complicated topics. In 2009, the median annual salary was $40,860, and the median hourly wage, $19.65. Interpreters and translators in the bottom 10 percent earned less than $22,810, while those in the top 10 percent earned more than $74,150”.

The first thing we ourselves might want to weigh is that these figures are not figures culled from the records of the United States Department of Labor, which statistically-speaking are much more accurate and reliable than those obtained from industry sources.  Rather, these figures are taken from ATA’s own bi-yearly compensation surveys, which for the most part are participated in by ATA members.  The response rate was 11.2% according to the 2007 ATA compensation survey report. Excluded from the survey are the association’s members who are not U.S. residents.  Therefore, the figures given above are based upon the responses of some 1000 translators and/or interpreters, utilizing survey methods that do not provide for the usual ±3% or ±4% accuracy margin. 

But let us say just for the sake of the argument (and also to stop some readers from complaining that I am being terribly unfair to the poor American Translators Association) that these figures do fall within the acceptable ±3% or ±4% accuracy goal.  Yes, let’s say that.  (See how fair I am being to the poor ATA?)  OK, now I would like all of you not to lose sight of the fact that the figures given by the ATA and published in “U.S. News & World Report” are in U.S. dollars and apply to translators (now more than ever CAT workers or operators) living in the United States.  Clearly, for those of you who reading this article from the comfort of your homes or offices in Mumbai  or Novosibirsk or Kobrin or Algiers or Cairo or Abidjan or Lima or Buenos Aires or Punta Arenas, a figure such as US$40,860 per annum may look like a king’s ransom.  Indeed, in Mumbai, if such an annual salary is not kingly, it is certainly princely.

But if you happen to live in the U.S. or even western Europe, then US$40,860 (ca. € 29,200) is not exactly the kind of “bread” that will allow you to buy a first-class airline ticket to see your family over the Christmas holiday.  Come to think of it, it may not even buy you a ticket in the plane’s cargo bay.    You see, in today’s United States, earnings of US$40,860 would not be bad at all if you were living rent-free at momma’s house, contributed about $25 a week towards meals and had free use of daddy’s car to run your errands.  And if you were still under momma’s and daddy’s medical insurance policy (with momma and daddy paying the premiums), your US$40,860 might even entitle you a nice 10-day vacation somewhere in the Caribbean. 

However, let’s change the venue and conditions somewhat.  Suppose you are a translator earning US$40,860, have a spouse who earns the same amount, live in an urban area and have two children.  And let’s say for the sake of the argument that this US$40,860 is the gross amount that you earn as a self-employed freelance translator (or CAT operator), from which first you must deduct your business expenses.  Well, the first thing you better hope for is that your spouse’s annual earnings of US$40,860 comes from salary or wages, and that his or her employment also provides full family healthcare benefits (including dental care.  And if those two children are of college (university) age, you better start hoping that your parents or your spouse’s parents are amenable to opening their pocketbooks.

In other words, gross annual earnings in the United States of US$40,860 from self-employed translation activities is plain chopped liver (to use an American idiom).  For “U.S. News & World Report” to categorize translation (and/or interpreting) with its mean annual earnings of US$40,860 as one of the best 50 jobs in the U.S. in 2011 is understandable; for their purpose is to sell magazines, and if the purveying of (economic) nonsense sells magazines, well that’s part of the business of selling magazines.  But for an organization such as the American Translators Association that purports itself to represent an industry or profession and further purports itself to be the guardian and protector of translator good & welfare in the United States, to be in direct partnership with this shameless purveying of economic nonsense, i.e., that translation with its median hourly wage of US$19.65 somehow constitutes a “great occupation” or a “growth occupation”  or “the occupation of the future” (particularly when one takes into account that a common rural postal carrier in the U.S. earns an hourly wage of US$18.50 (plus a package of social benefits, including healthcare insurance), is to me cynicism  at  best, and something indescribable at worst.  Indeed, while I had no problems categorizing ATA’s so-called “certification program”  as intellectual fraud, I do find myself groping for a descriptor for this latest act of sheer unadulterated cynicism and snake oil salesmanship by one of the largest if not the largest translator-interpreter organizations in the world.     

And then there is another side to this statistical story which neither the ATA nor “U.S. News & World Report” (fed with its ATA-generated statistics) appear disposed to mentioning.  And that side is about translators-interpreters who may earn as much as US$240,000 per annum. Of course, one of the qualifications for earning that princely amount is that you have to speak and/or read and/or write one of what former ATA President Marian S. Greenfield unabashedly called “the languages of terrorism”.  And you have to be prepared to earn those dollars in one of those countries where one of those “languages of terrorism” (Greenfield’s words, not mine) are spoken and/or written and/or read.  And you equally have to be prepared to sacrifice perhaps one limb or maybe two, or perhaps an eye (no big deal, you’ll still have another one), or be prepared to shoulder a “slight” concussion that will make your brains resemble scrambled eggs.  And of course, you’ll have to find the right government contractor to whom to render your translation and/or interpreting services. 

Now even at the risk of boring some of you, permit me to add just a couple more wee elements into this cauldron of statistics…er, excuse me, “stats” (I really don’t want any of you to think that I am doddering or incontinent or on my way to The Big Dictionary in the Sky):  Unless you have been living in a deep cave for the past five years (at the very least), you are aware, perhaps even painfully that more and more translation is being done by or with the assistance of computers.  Indeed, Mr.Kirti Vashee, who claims expertise in this area, said precisely that in a comment made here to one of Rosene Zaros’ articles.  And Mr. Vashee additionally implied that this volume of computerized or computer-assisted translation will increase.  And all of the translation technologists and machine-translation gurus and gurettes are telling us that in the coming decade, the bulk of translation activities will be one way or the other in the hands of computers.  And since machine translation and any of its mutations are far more “cost-effective” (in plain language, cheaper) than human translation, do not we wish to ask the question as to what kind of effect will this have on those median annual earnings of US$40,860?  Patience. Read on:

Many of those same translation technologists and machine-translation gurus and gurettes are also telling us, nay, firmly predicting that the future of translators (and CAT operators) is in post-editing et le cas échéant pre-editing of the detritus issuing from the translation machine.  Good.  At least there is a rosy spot in our employment picture.  We will go from that not-so-prestigious job title of translator (or interpreter) or the absolutely non-prestigious job title of CAT worker, to the prestigious-sounding title of translation editor (with emphasis on editor).  I like that.  “I’m a translation editor”.  Hell, it sounds almost as good as “I’m a newspaper or magazine editor”.  But what about the earnings figures from all that pre- and post-editing?

One of my correspondents recently wrote the following to me:

“I recently attended an online Webinar entitled "Machine Translation in the Real World:  A Dell Case Study".   The webinar was conducted by an outfit called Applied Language Solutions and the "Dell Case Study" seemed to be nothing more than an attempt to give credibility to Applied Language Solutions.   Part of the session included a history of machine translation from its early beginnings to what presenters described is now approaching a state-of-the-art application.  They did admit that while machine translation is not yet fully a state-of-the-art application, its status as such is clearly on the horizon. Some of the attendees did ask about price; e.g., how would machine translation with post-editing compare to having a document translated by a human translator?  The answer was that, initially it would be about 50% of the cost of human translation but that would decrease as more and more corpora (another buzz word) was/were (not sure if its singular or plural) added.  The corpora (at least 1,000,000 words) consists of translation memories, glossaries and documents that have already been translated.  .

While the presenters acknowledged that MT is not really suitable for idiomatic and idiosyncratic texts, they clearly downplayed those aspects of translation and stressed the fact that post-edited machine translation offers major cost savings.  Therefore, there would be a discernible need for post-editors.  And where would these post-editors come from?  The response was that they might come from the rank and file translators whose responsibilities would include the insertion of nuance into the final translation, along with any cultural implications, idiosyncratic language, etc.  In the overall scheme, these post-editors would function at a relatively low level”.

Since the cost prediction for the machine translation process is 50% less than that of human translation, one might want to ask whether this new corps of post-editors will receive 50% less for their endeavors than what today’s translator (in the U.S. … to keep us focused on the “U.S. News & World Report” crystal ball) receives.  Clearly, that 50% would be 50% of what are already depressed prices for human translation services, particularly in North America and western Europe.  That is not an exactly a rosy picture.  Yet even in the light of (1) the continued replacement of human translation by robotic translation and (2) a current median annual income of just under US$41,000 in the United States, an organization like the American Translators Association is proudly and loudly telling not just its members but the public-at-large that translation is not just coming up roses, it is roses.  According to the snake oil marketers at ATA, translation is one of the most economically-attractive occupations and one with a truly promising future.  I seem to smell the pungent aroma not just of bovine and equine excrement, but also of porcine excrement.

Welcome to 2011 and a Very Happy New Year!


  1. There is one small problem with all these rosy predictions about machine translations. MT produces still the same garbage that it produced 10 or 20 years ago. I tested Google Translate on a Japanese patent and wrote a blog about it. The result of Google Translate was pathetic. This is not something that can be used by inventors or patent lawyers.

    The good news is that machine translation does not work and never will.

    Happy New Year!

    Steve Vitek

  2. Steve - Even in the light of unimpeachable translation facts, those like you who question the High Priests of Translation Technology are labeled "Luddites" or "Translators of Another Time" at best, and some other descriptors at worst.

    I also congratulate you on PatentTranslator having been included in ATA's list of blogs. The heretical and savage Translation Commentator has not been accorded such "honors".

  3. "The heretical and savage Translation Commentator has not been accorded such "honors"."

    That's why I linked to it on my blog.
    It will be picked eventually by other blogs, I am sure of that.

    Gutta cavat lapidem non vi, sed saepe cadendo.

    Best regards,

    Steve Vitek

  4. I'm the CEO of Applied Language Solutions and thought it would be worth clarifying a couple of points.

    In a Post Editor/Translation Editor role productivity is at least doubled, if not more and therefore earnings are often better doing this type of work than traditional translation work.

    We build statistical machine translation models, specific to a client and the feedback on quality we get from this is extremely positive from translators. This is not a general engine such as Google and is very specific to the content being translated. It is not right for all types of translation and may never be. It is however here to stay and it will most definitely grow exponentially over the coming years.

    There is genuine opportunity for translators who want extra work and extra earnings from this. Similar to CAT tools it won't be for everyone and that is perfectly understandable, it is however a genuine opportunity for translators who want to embrace this as part of their overall offering.

    For any of your readers interested in the full context of the Webinar it is available here to view:

    Gavin Wheeldon

  5. You write, “In a Post Editor/Translation Editor role productivity is at least doubled, if not more and therefore earnings are often better doing this type of work than traditional translation work.”

    Sir, I do not know the make-up of the readership of Translation Commentator in terms of either its overall experience and sophistication about the industry and its economics, although I do know that the publisher of this blog and journal is a woman of both experience and economic sophistication with respect to this industry, and I am sure she can speak for herself as regards your above-transcribed comment. As to myself, although I no longer have as fancy a title as CEO and my retired military rank is not of three letters like GEN or COL, I do hope that you will provide me (despite my rank-less business and/or relatively low-rank military status) and the readers of this august publication with a wee bit of an explanation as to how a person’s productivity can be doubled doing post-translation editing. And while you are in that territory, perhaps you would deign to explain to us unwashed technological masses how even with all of this magic productivity involving a few dottings of “i’s” and crossing of “t’s” compensated at levels anywhere between one-half and less of current human translation fees, this new post-translation editor will earn more than he or she ever earned as a translator. Yes, could you please explain this, if not to me, then at least to the readership of Translation Commentator.

    And you also write, “There is genuine opportunity for translators who want extra work and extra earnings from this”.

    I truly hate to tell you this because it may come across that I am poking fun at your form of expression, which I really am not since I am more than aware that your field of expertise is technology and computer software and not language, linguistics and communication (which is the field of expertise of us silly, lowly, rank-less translators). Anyway, no fault of your own, sir, but that above-transcribed sentence has the following message, flavor, tone and nuance: “THIS IS A GREAT OPPORTUNITY FOR THOSE OF YOU WHO AFTER COMING HOME FROM YOUR JOB AT TACO BELL OR BURGER KING TO DO SOME EXTRA WORK AND MAKE SOME EXTRA MONEY”. Will your company of which you are the CEO be advertising on Craig’s List?

  6. Maybe a little math might do the trick and shed light on the question how a post-editor might earn more than a translator who stays away from computer-aided translations:
    Say I am an average manual translator (in output only, to be sure) and can deliver 2500 words per day of publication-worthy translations (leaving aside, for now, the fact that the 2500 words output/day will hardly provide a sufficient income these days).
    My earnings then will be 2500 times the agreed-upon per word rate, 2500x.
    As an average post-editor with some experience, I can deliver 5000 to 8000 words of publication-worthy translation. Even at one-half of the full translation rate, my earnings will be the same (at the low end)
    5000 times x/2=2500x or even more:
    8000 times x/2= 4000x.
    Of course, this is mere theory, because few of us are lucky enough to have 2500 words to translate day in and day out, year-round. So, even if I do prefer to translate without CAT tools, would it not be useful or even advisable to expand my horizons, learn about CAT tools and apply this in times when manual translation work is sparse? (Of course, if this never happens to me, and I am fully booked, never having to worry about idle times, then more power to me. But in that case all these discussions are moot, no?)
    Anyway, back to my milk-maid's math: If I am unlucky enough to only have a manual translation project of about 1500 words, then as an exclusively manual translator my income is 1500x. And I have the afternoon off, like it or not. But if I also have post-editing work, I can provide another 2500 words to a client, thus ending the day with
    1500x + 2500x/2 = 2750x.
    Not only have I saved the day, but I actually come out ahead.
    In this light, does it not seem that expanding one's capabilities, adding new skills, is prudent? In an ideal world, I would have my order books filled with manual translations that will provide me with an acceptable income. But since this world is anything but ideal, this is unlikely. Post-editing, even at lower rates than manual translation, can be an opportunity for translators indeed (not so much for fast food workers, though).